I started with a ship full of max-level veterans, and I'm just a complete noob.

Three days later, Fang Xudong placed the first volume of original device records in the center of the conference table at the Institute of Physics. Huang Kun glanced at the list of attendees and gestured for him to start directly from the test data.

Wu Hanzhang sat to Jiang Ming's left, with the preliminary research plan on his lap. The word "silicon" on the cover and inside page was still written in pencil, and the question mark was still there.

Fang Xudong first showed the frequency distribution of two batches of germanium double-diffused triodes of 502, and then put the base region profile photo, leakage current curve and Xie Xide's sample number into the same table.

The base region width of the first batch of tubes is concentrated in the range of 0.6 to 0.7 micrometers, the surface state density is relatively high, and the cutoff frequency mainly falls in the range of 120 to 150 MHz.

The second batch uses acid pickling, low-vacuum annealing and silicon oxide encapsulation, which reduces the surface state density by about 50% and raises the cutoff frequency to 160 to 180 MHz.

Xie Xide took over the conversation and laid out the photoconductivity spectrum of the germanium sample. The defect peak in the middle of the band gap circled with a red pencil fell under the same processing conditions as the high leakage current region of the 502 device.

"Based on minority carrier lifetime and surface recombination rate, the surface state drags down the cutoff frequency by about 10% to 15%, and the changes before and after sample processing support this range."

She placed another set of silicon surface state curves above the germanium curves, only showing the measurement conditions to the audience for the time being, leaving the material differences to be discussed in the third topic.

Wang Shouwu then opened his gray-blue briefcase and took out the test booklet from the semiconductor research group. The data he brought showed a similar upper frequency limit to that of the 502, which used different germanium materials and different junction formation methods.

The highest sample tube reached 190 MHz, while the main distribution was still between 150 and 180 MHz. Changes in process batches could only cause slight shifts in the curve.

Lin Lanying recalculated the two sets of data according to purity, resistivity and junction width, and after deducting the contributions of surface states and base width, she marked the estimated range of 250 to 300 MHz on the graph paper.

The first two topics took an hour and a half. The three parties reached a consensus on the quantitative contribution of 10% to 15% of the surface states and also acknowledged that there is still room for improvement in the existing germanium process.

Huang Kun turned to the last page of the agenda, his pen landing next to the third item.

"Now that we're discussing the device roadmap for the 24 gigahertz spectrum, who wants to speak first?"

Wang Shouwu moved the enamel mug to his side, pulled out a page from his thin leather notebook from the Purdue era, and copied it down. The edges of the paper were already yellowed. It contained the latest report on alloy germanium transistors from American laboratories.

"The alloy method for germanium transistors has already reached 502 MHz. If the crystal orientation control, surface treatment, and high-purity materials of the research group can be combined, it is realistic to achieve the same level."

He used a pen to point out the junction structure in the data, explaining the alloy junction, base width, and package lead inductance item by item, his tone still retaining the confidence of a materials researcher in germanium.

"Germanium still has value in intermediate frequency communication, pulse switching and replacement of existing equipment. A 24 GHz preamplifier can supplement cavity oscillators, parametric amplifiers or other devices, and the material route does not need to be cut off all at once."

After listening to everyone's speech, Jiang Ming moved his chair back, picked up the chalk, and walked to the blackboard.

He erased the curves left by the first two issues and drew a coordinate system on the left, middle and right of the blackboard, with the horizontal axis labeled with the wave vector k and the vertical axis labeled with the energy E.

The space between the conduction band and valence band on the left is labeled Germanium, 0.67 electron volts; the middle label is silicon, 1.12 electron volts; and the right label is gallium arsenide, 1.42 electron volts.

Wang Shouwu was still flipping through the alloy law data when he saw the third picture appear. His hand stopped at the corner of the paper. He had originally guessed that Jiang Ming would advocate for silicon, but gallium arsenide was beyond the scope of the current discussion.

Jiang Ming has not yet started the gallium arsenide process; he has only written a set of room temperature intrinsic carrier concentrations for each of the three band diagrams.

Germanium is approximately 2.4 x 10 to the power of 13 per cubic centimeter, silicon is approximately 1.5 x 10 to the power of 10, and gallium arsenide is approximately 2.1 x 10 to the power of 6.

When the chalk reached the last index, the only sound in the meeting room was the scratching of paper on the table. Seven orders of magnitude differences were laid out side by side on the same blackboard.

"These three sets of numbers are determined by the atomic electronic structure and the lattice periodic potential. Purity, equipment precision, and furnace investment can only improve the process, but cannot change the band gap of the material itself."

Jiang Ming plotted the changes in intrinsic carriers after temperature rise under the germanium figure, and then connected it to leakage current, breakdown voltage, junction capacitance and operating temperature range.

"We can push germanium transistors to 350 MHz, and we may continue to push them upwards through narrower base regions and better packaging. The preliminary research target can be from 100 MHz to more than 1 GHz, and material constraints will become more and more concentrated."

Wang Shouwu stared at the three sets of numbers, his mind still recalculating the relationship between temperature, leakage current and minority carrier lifetime. He was familiar with each formula and had read the corresponding data, but in the past they were scattered in different papers and process problems.

Now, with the diagram, the measured values ​​of both devices, and the 24 gigahertz target all on the same blackboard, the question mark on the last page of his notebook re-enters his field of vision.

Xie Xide first picked up his silicon sample curve and placed the defect peaks in the middle of the bandgap of germanium and silicon side by side on the table.

"The band gap determines the relative strength of the pinning of the surface states to the Fermi level. In the range of 0.67 electron volts, the proportion of existing defect levels is too high."

She traced the silicon curve with a pencil, stopping at the low-value area after oxidation.

"The 1.12 electron volts provides more leeway for interface processing, and the interface between silicon and silicon oxide is more stable. From the perspective of active devices, silicon has the potential to continuously shrink the junction area and reduce leakage current."

After listening to this, Huang Kun wrote down the band gap and intrinsic carrier concentration of the three materials in his hardcover notebook, and then drew a horizontal line below them.

Below the horizontal line is still a blank space. He still needs to determine whether the material direction can be connected with the existing equipment capabilities. Physical advantages alone cannot support process decisions.

Lin Lanying opened the physical properties data and wrote down 937 degrees and 1420 degrees next to germanium and silicon respectively, and then marked the word "furnace" on the margin.

The word reminded her of the quartz tubes, heating elements, and existing temperature zones of Furnace No. 3. Silicon conversion meant that the entire equipment would have to withstand higher loads, and the physical hope would immediately turn into processing and supply pressure.

Wang Shouwu brought the enamel cup to his lips; the tea inside had already gone cold. He then placed the cup back on the table, the bottom of which struck the wooden table with a dull thud.

"You mean all the work we did on germanium—purification, refining, single crystal formation, dicing, and diffusion—was wasted?"

Jiang Ming glanced at the thin notebook beside him, knowing full well that this sentence referred to far more than just a few batches of sample tubes; the semiconductor research group had invested furnace time, personnel, and judgment all in it.

He now has to answer far more than just the selection of materials; he also needs to prove what the hard work he has put in over the past few months has actually yielded.